All the World to Gain
At the end of the last post we arrived at a conception of knowledge called Reliabilism. According to this conception, the necessary and sufficient conditions for knowledge are (i) that the knower have a belief, (ii) that the belief is true, and (iii), that the belief is produced by means that are reliable in the circumstances. Recall the skeptical challenge:
Imagine, for a moment, that the world is not as it seems to be. Imagine that the experiences which are so formative in your conception of the world are not the product of various objects they superficially indicate, but are instead the product of a carefully designed computer programme. You are a brain in a vat, and your experiences are stimulated by electrical impulses, sent from a supercomputer via electrodes attached to various parts of your brain. If that were the case, how could you know it? But perhaps it is the case. How, then, can one claim to know anything about the external world?
We can now reply. Our beliefs about the external world, such as the belief that ‘I have hands’, count as knowledge iff1 that belief is true and it is reliably produced. In ordinary, rather than skeptical circumstances, this will be the case. It will be true that I have hands, and this belief will come by way of my senses, chiefly from my visual apparatus, but also from my tactile faculties – reliable processes all. And so skeptical hypotheses, so long as they are false, fail to threaten my ordinary claims to knowledge. Recall now our two principles, which the skeptic used to make his case:
EP: S knows that p on the basis of evidence e, only if e entails p.
PDC: If S knows that p, and S knows that p entails q, then if S believes that q, S also knows that q.
It was argued previously that if the Entailment Principle were true, then the possibility of a skeptical alternative to our commonsense conclusions about the world would deprive us of knowledge. Given the Reliabilist conception of knowledge, what can we say about this principle? I think it depends upon how it is interpreted. If the EP is read to imply that e is the exclusive basis of knowledge that p, then it may well be true, for the antecedent describes something which is impossible according to Reliabilism2 – that is, that S knows that p only on the basis of evidence e. As we have seen, knowledge requires more than evidence, though evidence may enter into the description of a reliable process. On the other hand, if the EP is read to imply that knowledge grounded in evidence e is impossible unless e entails p, then the principle is straightforwardly false. As it happens, we quite often have knowledge without entailing evidence, because entailing evidence is not required to have a reliably formed true belief. Either way, the Entailment Principle is not an insuperable barrier to knowledge of the external world.
The other way to skepticism was via the Principle of Deductive Closure. With this principle in hand, the skeptic argued that, unless we were prepared to make the audacious claim to know the falsity of a skeptical possibility, we could not claim to know any of the mundane facts about the world which seem so obvious. Is this a conclusive argument for skepticism? At first blush it looks as though we are in a worse position as a result of our analysis of knowledge. That is because the deductive reasoning which the principle expresses looks to be a reliable process, and we must grant this pivot of the skeptic’s argument.
We can, however, disagree with the skeptic over the principle’s implications. It may be sagely observed that, though the skeptic reasons that our lack of knowledge concerning the truth of a skeptical possibility requires us to give up our everyday knowledge claims, one could just as well apply the principle in reverse and argue that these knowledge claims deductively entitle us to knowledge that the skeptic’s hypothesis is false. One person’s modus tollens is another person’s modus ponens, so why not reply in kind? Whilst this reply is less than decisive, it does show that merely presenting a logical principle like the PDC is not adequate to adjudicate the dispute.
Still, one feels that the skeptic presents us with a puzzle. It certainly feels as though the skeptic has the upper hand in making his case, and that we should acquiesce rather than he. So why should the dialectic tend to favor the skeptic’s destructive speculations, instead of commonsense? The answer seems to lie in implicit norms which govern what it is appropriate to assert in various conversational contexts. Begin with the idea that, in making a knowledge claim, we have at least one fellow conversant in mind, and as such our efforts are directed toward convincing that person. Convincing the other person involves arguing from a common ground of shared assumptions, and showing that on the basis of such assumptions, some further proposition is to be believed. So, we have a goal (persuasion), and a rule which must be adhered to for the goal to be attained (relying only on common ground to make one’s case). We therefore have a norm which governs this kind of discourse: only assert that which is supported by the common ground shared by you and your fellow conversant.
Now, take any ordinary knowledge claim, offered by one non-skeptic to another – let’s use “I know that’s Notre Dame”, said pointing to a Gothic structure in Paris. If Reliabilism is true, the convincee will implicitly assess the claim in accordance with the three Reliabilist conditions. Presuming that she agrees the building is Notre Dame, and presuming also that she has no reason to doubt the sincerity of the claimant, she can take it that the truth and belief conditions are met. And if she inquires after the grounds for the belief, and is told that the claimant has come to the belief via a combination of eyesight, memory, and testimony (say, that the claimant has read books about Notre Dame cathedral), she will grant him knowledge, as she shares with the claimant the assumptions about the reliability of his sources.
But contrast this with the retort to the skeptic, perhaps that one knows ‘I have hands’, and therefore can dispense with the skeptical hypothesis. Whereas the claim made in our previous context could be supported by appeal to common ground, any simple knowledge claim made in response to the skeptic cannot be supported on common ground, for the skeptic refuses to yield on any non-necessary matter. But this rhetorical advantage, once understood, obviously does not amount to any kind of logical demonstration. Those who believe that they have knowledge of the external world are not rationally required to give up these beliefs simply because, if they are fool enough to think that they can convince the skeptic, they will be discursively required to feign the ignorance of their interlocutor.
Indeed, it does not look as though the skeptical argument which employs the PDC has any logical traction at all. For, if Reliabilism is true, then having knowledge does not turn on the question of what we might be prepared to claim as knowledge, what we might have to justify. Having knowledge is primarily a matter of the external world’s being a certain way: of the truth of one’s belief and the reliability of processes whereby one comes to that belief. And so, unless a skeptical hypothesis is true, it seems that our ordinary knowledge claims are largely correct. I conclude, then, that neither of the skeptical arguments considered put our ordinary knowledge claims in jeopardy. The next post will consider whether the skeptic can reply from considerations of rationality instead.
1 The word ‘iff’ is a contraction of ‘if and only if’.
2The entailment principle is a conditional, and conditionals are only false in the case where the antecedent is true and the consequent is false. But if the antecedent is impossible, it can never be true, which means that the antecedent cannot be true whilst the consequent is false, and so the conditional as a whole can never be false.